| This dissertation explores wildfire dynamics. Chapters 2 and 3 are peer reviewed journal articles that present an understanding of three-dimensionality in grass fires and how it affects forward rate of spread (ROS) of the fire. In Chapter 2 the numerical model HIGRAD/FIRETEC was used to give arguments supporting that modeling wildfire in a two-dimensional vertical and stream-wise plane does not represent all of the physics that are required to determine a meaningful forward ROS. Chapter 2 inspired the work that makes up chapter 3. In chapter 3, HIGRAD/FIRETEC was again used, to determine the effect that ignition line length has on forward ROS. In both chapters, finger shaped structures were present in the combusting fuels, upstream of the fire front. The fingers correlated with counter-rotating vortex pairs in the gas-phase above them. It was also shown that increasing ignition line length does indeed increase forward ROS, an expected result supported by previous investigations. Results were presented that suggest physical reasons why a spreading grass fire develops flanks that move forward slower than the front of the fire.;Chapter 4 describes the gas phase in the planetary boundary layer (PBL), where fires and other phenomena occur. A muti-component gas phase model was derived that represents individual ideal gas species. The mass dependent nature of this model allows the individual species to have dynamic effects on the flow field. The multi-component model was then coupled to HIGRAD to explore three PBL scenarios. The purpose of the first case was to numerically spin-up a moist unstable PBL. The second case used the mixture model to look at a hypothetical scenario representative of the Las Conchas wildfire. In the second case, an idealized column of a gaseous mixture containing heat, dry air, water vapor, and fullerene was initialized over the topography where the Las Conchas fire occurred. The gas column represented an idealized fire plume. As predicted, the column collapsed under its own weight. However, parts of the column rose to higher elevations too. The third case was to use the results from the first case to model fugitive methane in an unstable PBL. In this case, a small amount of methane was fluxed into a grid cell on the bottom boundary of the spun-up, moist PBL from the first case. The evolution of the methane plume was and continues to be studied. Some preliminary results are presented where methane concentration is compared to data collected from a field experiment.;Chapter 5 is the conclusion. In this chapter, the results from chapters 2, 3, and 4 are summarized. An analytical fire spread model is proposed that ties chapters 2 and 3 to chapter 4. This spread model is a two-layer, multi-phase set of governing equations. It is assumed that the gas phase is a thin layer relative to the horizontal. This thin layer contains laminar, diffusion flames. The vertical distance that the flames remain laminar and diffusive determines the thickness of the layer. In other words, the vertical distance from the ground to the transition point to turbulent flames, determines the thickness of the layer. |
